1
0
Fork 0
mirror of https://github.com/moby/moby.git synced 2022-11-09 12:21:53 -05:00
moby--moby/volume/volume.go
Kir Kolyshkin 516010e92d Simplify/fix MkdirAll usage
This subtle bug keeps lurking in because error checking for `Mkdir()`
and `MkdirAll()` is slightly different wrt to `EEXIST`/`IsExist`:

 - for `Mkdir()`, `IsExist` error should (usually) be ignored
   (unless you want to make sure directory was not there before)
   as it means "the destination directory was already there"

 - for `MkdirAll()`, `IsExist` error should NEVER be ignored.

Mostly, this commit just removes ignoring the IsExist error, as it
should not be ignored.

Also, there are a couple of cases then IsExist is handled as
"directory already exist" which is wrong. As a result, some code
that never worked as intended is now removed.

NOTE that `idtools.MkdirAndChown()` behaves like `os.MkdirAll()`
rather than `os.Mkdir()` -- so its description is amended accordingly,
and its usage is handled as such (i.e. IsExist error is not ignored).

For more details, a quote from my runc commit 6f82d4b (July 2015):

    TL;DR: check for IsExist(err) after a failed MkdirAll() is both
    redundant and wrong -- so two reasons to remove it.

    Quoting MkdirAll documentation:

    > MkdirAll creates a directory named path, along with any necessary
    > parents, and returns nil, or else returns an error. If path
    > is already a directory, MkdirAll does nothing and returns nil.

    This means two things:

    1. If a directory to be created already exists, no error is
    returned.

    2. If the error returned is IsExist (EEXIST), it means there exists
    a non-directory with the same name as MkdirAll need to use for
    directory. Example: we want to MkdirAll("a/b"), but file "a"
    (or "a/b") already exists, so MkdirAll fails.

    The above is a theory, based on quoted documentation and my UNIX
    knowledge.

    3. In practice, though, current MkdirAll implementation [1] returns
    ENOTDIR in most of cases described in #2, with the exception when
    there is a race between MkdirAll and someone else creating the
    last component of MkdirAll argument as a file. In this very case
    MkdirAll() will indeed return EEXIST.

    Because of #1, IsExist check after MkdirAll is not needed.

    Because of #2 and #3, ignoring IsExist error is just plain wrong,
    as directory we require is not created. It's cleaner to report
    the error now.

    Note this error is all over the tree, I guess due to copy-paste,
    or trying to follow the same usage pattern as for Mkdir(),
    or some not quite correct examples on the Internet.

    [1] https://github.com/golang/go/blob/f9ed2f75/src/os/path.go

Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
2017-11-27 17:32:12 -08:00

231 lines
7.6 KiB
Go

package volume
import (
"fmt"
"os"
"path/filepath"
"syscall"
"time"
mounttypes "github.com/docker/docker/api/types/mount"
"github.com/docker/docker/pkg/idtools"
"github.com/docker/docker/pkg/stringid"
"github.com/opencontainers/selinux/go-selinux/label"
"github.com/pkg/errors"
)
// DefaultDriverName is the driver name used for the driver
// implemented in the local package.
const DefaultDriverName = "local"
// Scopes define if a volume has is cluster-wide (global) or local only.
// Scopes are returned by the volume driver when it is queried for capabilities and then set on a volume
const (
LocalScope = "local"
GlobalScope = "global"
)
// Driver is for creating and removing volumes.
type Driver interface {
// Name returns the name of the volume driver.
Name() string
// Create makes a new volume with the given name.
Create(name string, opts map[string]string) (Volume, error)
// Remove deletes the volume.
Remove(vol Volume) (err error)
// List lists all the volumes the driver has
List() ([]Volume, error)
// Get retrieves the volume with the requested name
Get(name string) (Volume, error)
// Scope returns the scope of the driver (e.g. `global` or `local`).
// Scope determines how the driver is handled at a cluster level
Scope() string
}
// Capability defines a set of capabilities that a driver is able to handle.
type Capability struct {
// Scope is the scope of the driver, `global` or `local`
// A `global` scope indicates that the driver manages volumes across the cluster
// A `local` scope indicates that the driver only manages volumes resources local to the host
// Scope is declared by the driver
Scope string
}
// Volume is a place to store data. It is backed by a specific driver, and can be mounted.
type Volume interface {
// Name returns the name of the volume
Name() string
// DriverName returns the name of the driver which owns this volume.
DriverName() string
// Path returns the absolute path to the volume.
Path() string
// Mount mounts the volume and returns the absolute path to
// where it can be consumed.
Mount(id string) (string, error)
// Unmount unmounts the volume when it is no longer in use.
Unmount(id string) error
// CreatedAt returns Volume Creation time
CreatedAt() (time.Time, error)
// Status returns low-level status information about a volume
Status() map[string]interface{}
}
// DetailedVolume wraps a Volume with user-defined labels, options, and cluster scope (e.g., `local` or `global`)
type DetailedVolume interface {
Labels() map[string]string
Options() map[string]string
Scope() string
Volume
}
// MountPoint is the intersection point between a volume and a container. It
// specifies which volume is to be used and where inside a container it should
// be mounted.
type MountPoint struct {
// Source is the source path of the mount.
// E.g. `mount --bind /foo /bar`, `/foo` is the `Source`.
Source string
// Destination is the path relative to the container root (`/`) to the mount point
// It is where the `Source` is mounted to
Destination string
// RW is set to true when the mountpoint should be mounted as read-write
RW bool
// Name is the name reference to the underlying data defined by `Source`
// e.g., the volume name
Name string
// Driver is the volume driver used to create the volume (if it is a volume)
Driver string
// Type of mount to use, see `Type<foo>` definitions in github.com/docker/docker/api/types/mount
Type mounttypes.Type `json:",omitempty"`
// Volume is the volume providing data to this mountpoint.
// This is nil unless `Type` is set to `TypeVolume`
Volume Volume `json:"-"`
// Mode is the comma separated list of options supplied by the user when creating
// the bind/volume mount.
// Note Mode is not used on Windows
Mode string `json:"Relabel,omitempty"` // Originally field was `Relabel`"
// Propagation describes how the mounts are propagated from the host into the
// mount point, and vice-versa.
// See https://www.kernel.org/doc/Documentation/filesystems/sharedsubtree.txt
// Note Propagation is not used on Windows
Propagation mounttypes.Propagation `json:",omitempty"` // Mount propagation string
// Specifies if data should be copied from the container before the first mount
// Use a pointer here so we can tell if the user set this value explicitly
// This allows us to error out when the user explicitly enabled copy but we can't copy due to the volume being populated
CopyData bool `json:"-"`
// ID is the opaque ID used to pass to the volume driver.
// This should be set by calls to `Mount` and unset by calls to `Unmount`
ID string `json:",omitempty"`
// Sepc is a copy of the API request that created this mount.
Spec mounttypes.Mount
// Track usage of this mountpoint
// Specifically needed for containers which are running and calls to `docker cp`
// because both these actions require mounting the volumes.
active int
}
// Cleanup frees resources used by the mountpoint
func (m *MountPoint) Cleanup() error {
if m.Volume == nil || m.ID == "" {
return nil
}
if err := m.Volume.Unmount(m.ID); err != nil {
return errors.Wrapf(err, "error unmounting volume %s", m.Volume.Name())
}
m.active--
if m.active == 0 {
m.ID = ""
}
return nil
}
// Setup sets up a mount point by either mounting the volume if it is
// configured, or creating the source directory if supplied.
// The, optional, checkFun parameter allows doing additional checking
// before creating the source directory on the host.
func (m *MountPoint) Setup(mountLabel string, rootIDs idtools.IDPair, checkFun func(m *MountPoint) error) (path string, err error) {
defer func() {
if err != nil || !label.RelabelNeeded(m.Mode) {
return
}
var sourcePath string
sourcePath, err = filepath.EvalSymlinks(m.Source)
if err != nil {
path = ""
err = errors.Wrapf(err, "error evaluating symlinks from mount source %q", m.Source)
return
}
err = label.Relabel(sourcePath, mountLabel, label.IsShared(m.Mode))
if err == syscall.ENOTSUP {
err = nil
}
if err != nil {
path = ""
err = errors.Wrapf(err, "error setting label on mount source '%s'", sourcePath)
}
}()
if m.Volume != nil {
id := m.ID
if id == "" {
id = stringid.GenerateNonCryptoID()
}
path, err := m.Volume.Mount(id)
if err != nil {
return "", errors.Wrapf(err, "error while mounting volume '%s'", m.Source)
}
m.ID = id
m.active++
return path, nil
}
if len(m.Source) == 0 {
return "", fmt.Errorf("Unable to setup mount point, neither source nor volume defined")
}
if m.Type == mounttypes.TypeBind {
// Before creating the source directory on the host, invoke checkFun if it's not nil. One of
// the use case is to forbid creating the daemon socket as a directory if the daemon is in
// the process of shutting down.
if checkFun != nil {
if err := checkFun(m); err != nil {
return "", err
}
}
// idtools.MkdirAllNewAs() produces an error if m.Source exists and is a file (not a directory)
// also, makes sure that if the directory is created, the correct remapped rootUID/rootGID will own it
if err := idtools.MkdirAllAndChownNew(m.Source, 0755, rootIDs); err != nil {
if perr, ok := err.(*os.PathError); ok {
if perr.Err != syscall.ENOTDIR {
return "", errors.Wrapf(err, "error while creating mount source path '%s'", m.Source)
}
}
}
}
return m.Source, nil
}
// Path returns the path of a volume in a mount point.
func (m *MountPoint) Path() string {
if m.Volume != nil {
return m.Volume.Path()
}
return m.Source
}
func errInvalidMode(mode string) error {
return errors.Errorf("invalid mode: %v", mode)
}
func errInvalidSpec(spec string) error {
return errors.Errorf("invalid volume specification: '%s'", spec)
}